Using Machine Learning to Better Understand Environmental Policy Innovation – UROP Spring Symposium 2021

Using Machine Learning to Better Understand Environmental Policy Innovation

Rohan Bhargava

UROP Logo

Pronouns: He/Him/His

Research Mentor(s): Michael Lerner, Ph.D. Candidate
Research Mentor School/College/Department: Political Science, College of Literature, Science, and the Arts
Presentation Date: Thursday, April 22, 2021
Session: Session 1 (10am-10:50am)
Breakout Room: Room 3
Presenter: 4

Event Link

Abstract

Climate change and environmental protection are perhaps the most important issues of our generation, and government policies and laws are key ways to address them. In order to better understand which countries are leaders in passing environmental protection laws, how they pass them, and what types of laws they pass, it is necessary to comb through databases of hundreds of thousands of laws, something impossible for humans to do. Therefore, it is necessary to use text analysis to more efficiently analyze laws. This project seeks to find the best method of computer text analysis to most accurately label and analyze laws and policies. In order to find the best method we are comparing the accuracy rates of a supervised transfer-learning model trained on laws we manually labeled in Atlas.ti, other more traditional machine-learning models we are building, and models used by others in similar literature. If we are able to construct a model with a high accuracy rate, it will make the analysis of law and policies much more efficient, which can help in a variety of research contexts. In our context, it would help us most accurately analyze how environmental policies have been passed worldwide.

Authors: Michael Lerner, Rohan Bhargava
Research Method: Library/Archival/Internet Research

lsa logoum logo